Really New Products

What it is

Steve Hoeffler established the first half of the problem, validating in an initial study that consumers have greater uncertainty when estimating the usefulness of really new products than of incremental ones, and showing that they cope by using inferential techniques that standard preference measurement, such as conjoint analysis, does not capture. His test of remedies, incorporating mental simulation and analogies into an existing measurement technique, found that some methods enhance predictive accuracy and others hinder it. The honest reading is that the instrument is unreliable for this class of product rather than biased against it.

In effect

David Alexander, John G. Lynch and Qing Wang supplied the second half across four field studies, finding that consumers follow through less often on positive purchase intentions for really new products than for incrementally new ones, and that the decrement grows over time. Compared with buyers of incremental products, buyers of really new ones think less concretely about the circumstances of buying and using the thing, and are more poorly calibrated about what initial use will be like. So the commercial failure is not that novelty tests badly. It is that novelty tests fine and then does not convert.

What it does not say

It does not line up with the organisational evidence on how novel ideas are received, and nothing here should pretend that it does. The evaluation literature finds panels and judges marking novelty down at the point of decision, while the consumer literature finds stated intentions towards novelty running warm and then failing, which is a different problem with a different remedy. No study in reach takes a genuinely novel product concept through standard pre-launch testing and shows that the novelty itself depressed the score, so the consumer version of the bias against creativity is not established. And no magnitudes are given here, since the abstracts carry none and the full texts were not read.

---